I wouldn't call it guessing to leverage billions in R&D spent by a global auto manufacturer to use as a minimum baseline, I'd call that smart, and critical to the process for someone that doesn't have an army of engineers and Elon Musk's net worth. That said, they DO have a non-zero acceptable failure rate in a given time period, and the harder we beat on our rigs the more likely we are that we'll find ourselves part of that statistic. So, if we decide that a design that met Chryslers accepted failure rate in their design conditions returns an unacceptable failure rate in our conditions, how do we know how much material to add? We obviously can't reproduce their R&D, so what do we do?
Clearly, as the inspiration of this thread, there are those that just make it so heavy and overdesigned that it can't possibly fail. Sure, it's sloppy and inefficient, but they're probably making a conscious decision that the profit they're losing out on due to material waste doesn't justify the cost of the upfront testing and design work to do it right. I make no claims to that being a right or wrong approach, because (provided they have data in some form) it can be right or wrong depending on the cost-benefit of the up-front analysis and testing.
I consider designing without data to be guessing. Data can come in different forms, including experience. It's not going to be as precise as the army of engineers and billion dollar testing facilities but it's the best most low-volume offroad parts producers are going to be able to do.
I'm a mechanical engineer, mostly a fluids/thermo/heat transfer guy, and though I don't call myself a fabricator, I've done enough mechanical component design to have an eye for what's necessary for most things if I have the factory part to go off of. I have no idea how anybody would even start designing a part without any of that experience informing their design process; so much so that it's hard for me to believe that there would truly be any significant numbers of people like that in the industry. But I've been wrong before.